EMPATHY FOR PAIN. And its Neural Correlates. Running head: EMPATHY FOR PAIN 2

Transcription

1 Running head: EMPATHY FOR PAIN 2 EMPATHY FOR PAIN And its Neural Correlates Bachelor Degree Project in Cognitive Neuroscience Basic level 15 ECTS Spring term 2016 Emelie Löfstrand Supervisor: Judith Annett Examiner: Paavo Pylkkänen

2 Running head: EMPATHY FOR PAIN 2 Abstract The phenomenon of empathy has been fascinating laymen and scholars for centuries and has recently been an important subject for cognitive neuroscientific study. Empathy refers to the ability to understand and share others emotions and a characteristic of this ability is the capacity to empathize with others in pain. This review intends to examine and read up on the current state of the field of the neural and behavioral mechanisms associated with empathy for pain. The neural underpinnings of the first-hand experience of pain have been shown to be activated in a person observing the suffering individual, and this similarity in brain activity has been referred to as shared networks. This phenomenon plays an important role in the study of empathy. However, different factors have been shown to influence empathy for pain, such as age, gender, affective link between observer and sufferer, as well as phylogenetic similarity. This thesis discusses these differences, as well as atypical aspects affecting the empathic ability such as synaesthesia for pain, psychopathy and Asperger s disease. Further, empathy for pain can be modulated by the individual observing someone in pain. For example, caregivers often down-regulate their empathic response to patients in pain, possibly in order to focus on their treatment and assistance. Also, paying attention to harmful stimuli heightens the perception of pain; therefore, the painful experience can be less remarkable when focusing on something else. The effect of empathy from others directed to oneself when suffering is discussed, as well as the consistency and limitations of presented research. Keywords: empathy, pain, shared networks, pain matrix, empathy for pain

3 EMPATHY FOR PAIN 3 Table of Contents Introduction 5 Empathy 9 Neural Correlates of Empathy 9 Mirror neurons 9 The empathy circuit 11 Pain 13 Theories of Pain 13 Gate control theory of pain 15 Pain Pathways 17 Neural Correlates of Pain 18 Empathy for Pain 19 Neural Correlates of Empathy for Pain 19 Shared networks 21 Somatosensory system 22 Factors Influencing Empathy for Pain 23 Age 23 Gender 25 Empathy for non-human entities 26

4 EMPATHY FOR PAIN 4 Empathy among caregivers 27 Racial bias 28 Atypical Aspects of Empathy for Pain 30 Synaesthesia for pain 30 Psychopathy 31 Huntington s disease 32 Asperger s syndrome 33 Post-traumatic stress disorder 34 Regulation of Empathy for Pain 35 Empathy Affects Pain Perception 37 Conclusions 38 References 43

5 EMPATHY FOR PAIN 5 Introduction The term empathy derives from the Greek word empatheia (passion), composed of en (in) and pathos (feeling). It was introduced into the English language from the German word einfûhlung (feeling into), which was then referred to as resonance with works of art, but later on came to describe the resonance between humans (Singer & Klimecki, 2014). The phenomenon of empathy and its many facets have been fascinating laymen and scholars for centuries (Lamm & Majdandžić, 2015). In the late 20 th century, researchers began to study this phenomenon on a scientific level. (Singer & Klimecki, 2014) In order to explain empathy, we first need to clarify how it differs from other related states which also have to do with the sharing of emotions. Empathy is commonly referred to as the ability to understand or feel what another being is experiencing (e.g. Singer & Klimecki, 2014; Fan & Han, 2008; Mu, Fan, Mao, & Han, 2008). According to some researchers, the sharing of emotions is a conscious process in which the other individual is considered the source of the state or the feeling (Betti & Aglioti (2016). However, others support the view that empathy is a multifaceted construct, consisting of several sub-constructs (Preston & de Waal, 2002). Empathy refers to the ability to understand and share others emotional states (Mu et al., 2008). For example, if someone is happy we get happy too, and if someone suffers from pain, we suffer too. Importantly, one does not confuse the feelings of another with one s own feelings. Empathy is not to be confused with emotion contagion, which is a state in which feelings are shared, but the self-other distinction is not present. Emotion contagion is commonly present in babies, where this distinction is not yet developed (Singer & Klimecki, 2014). It is assumed that emotion contagion is the foundation of empathy, and that its function is on the one hand on a phylogenetic level to provide a link between humans and other

6 EMPATHY FOR PAIN 6 species, and on the other hand on an ontogenetic level between adults and children (de Wall, 2008). When someone experiences negative emotions, the observer often experiences compassion, or more commonly referred to as sympathy. In this state, one does not share the feeling of suffering but rather experiences a warm and caring feeling of concern. This is accompanied by a motivation to help the other, a so called prosocial motivation (Singer & Klimecki, 2014). In sympathy, one feels for and not with the other. Another reaction to the suffering of others is empathic/personal distress. In this case, the observer perceives the suffering as self-oriented, and reacts with aversion to the situation. The observer has a strong desire to withdraw from the situation in order to avoid this negative feeling (Singer & Klimecki, 2014). As in the state of sympathy or empathic distress, empathy does not concern only negative feelings. On the other hand, empathy can occur regardless of the valence of the feelings, either positive or negative (Singer & Klimecki, 2014) The ability to experience empathy is not unique to humans; it seems to derive from antecedents in other species as well as rodents (Panksepp & Panksepp, 2013). There is also compelling evidence that other animals than humans possess the ability to experience sympathy; for example, some animals consolidate each other when losing a fight (de Waal, 2008). It is of great social importance to possess the ability to feel empathy for others. If we encounter a person who seems to lack this ability, they are usually considered to be either socially impaired or socially deviant (Bruneau, Jacoby, & Saxe, 2015). The ability to experience empathy is one of our most fundamental skills. It requires both emotional and cognitive functioning in the sense that one needs to be able to understand the thought and feelings of another, as well as take part in their emotions (Rameson, Morelli, & Lieberman,

7 EMPATHY FOR PAIN ). Therefore, many scientists differentiate between affective, or emotional empathy, and cognitive empathy (Ruckmann et al., 2015). Affective empathy refers to the sharing of feelings with others whereas cognitive empathy is the ability to understand others emotions, which is strongly connected to theory of mind (Eres & Molenberghs, 2013). For example, affective empathy is most often associated with activity in the insula, whereas cognitive empathy is, on the other hand, connected to activity in the midcingulate cortex and the adjacent dorsomedial prefrontal cortex (MCC/dmPFC). These findings provide validation for the hypothesis that empathy consists of several components. In other words, affective and cognitive empathy correlate with different neural and structural activities (Eres, Decety, Louis, & Molenberghs, 2015). Affective empathy is typically referred to as the conscious experience of others emotional states, which requires a self-other distinction, and an understanding of the origin of the emotion. In other words, this component of empathy reflects one s subjective experience of other beings emotions. Affective empathy differs from emotion contagion and mimicry, which are automatic responses and do notnecessarily require the distinction between self and other. It is also different from sympathy and empathic concern since they don t have to contain the sharing of emotions. Affective empathy has, on the other hand, been thought of as an umbrella term encompassing emotion contagion, mimicry, sympathy and empathic concern (Eres et al., 2015). Cognitive empathy refers to the ability to understand others motivation (Decety, 2011). Some researchers refer to cognitive empathy as Theory of Mind (ToM) (Mazza et al., 2015), although others make a distinction between empathy and ToM. Kanske, Böckler, Trautwein and Singer (2015) investigated a novel functional magnetic resonance imaging (fmri) paradigm called EmptaTom in order to reveal clearly separable neural networks for empathy and ToM. For the experience of empathy, there seemed to be a network

8 EMPATHY FOR PAIN 8 in the anterior insula, whereas for ToM, a network was found in the ventral temporoparietal junction. EmptaTom allows separation of the affective and cognitive routes to understanding others (Kanske et al., 2015). The understanding and experience of another person s pain, as well as other emotional states, is a characteristic of empathy. Many studies on empathy have been carried out with specific focus on pain. The brain areas involved with the direct experience of pain are also involved with empathy for the pain of others (Decety, Michalska, & Akitsuki, 2008) which makes the study of pain important for the understanding of empathy. Pain serves a function from an evolutionary perspective because it warns the subject and, in turn, the surrounding beings. The behavioral expressions as well as empathy for pain are necessary for us to be able to help and give care to the suffering individual (Craig, 2004). Facing a person that experiences pain can give rise to a variety of responses, ranging from ignoring to helping (Goubert et al., 2005). Usually, the sharing of the feeling of pain is an automatic response, although behavioral responses are distinguished by cognitive factors such as perspective taking, and emotion regulation such as motivation (Eres & Molenberghs, 2013). In this review, there will be a presentation of the similarities and differences between the direct experience of pain and empathy for another person s pain. The aim is to explore what happens in the brain of a person who observes another person suffering from pain in contrast to the direct experience of pain. First there will be a brief overview of the phenomenon of empathy and its neural correlates. Then there will be an overview of the physiology of pain and the different pathways and neural structures involved with painful experience. This section will provide an understanding of the direct experience of pain, from peripheral stimuli to the fundamental neural processes giving rise to a painful experience. Thereafter, the essay will focus specifically on empathy for pain. Here prominent research in

9 EMPATHY FOR PAIN 9 this field will be presented, including the disparity of empathy for pain in different neural deficits, genders and ages. Moreover, this section will also examine the factors which underlie how empathy for pain can be altered and regulated. Neural Correlates of Empathy Empathy A growing number of studies indicates that understanding and sharing of emotions of others depends on a recruitment of the same neural structures both associated with our own experience and when observing others experience the same thing (Rameson & Lieberman., 2009). The simulation theory of empathy suggests that our understanding of another being s emotions and thought is gained when we use our own minds as a model (Rameson & Lieberman, 2009) and the discovery of mirror neurons and shared networks have been considered to support this notion (Gallese & Goldman, 1998). In the following, the phenomenon of mirror neurons will be presented as well as an overview on its current debate. Mirror neurons. In 1992, an interesting discovery was made that showed that specific neurons in a monkey s brain fire when the monkey reaches for objects, as well as when it observes someone else perform the same action. This phenomenon was revealed by recording specific cells with electrodes inserted into the brain of the monkey. The monkey was then presented with a peanut and when it reached for this, the recorded cell fired. In another experiment, the monkey observed the experimenter reach for a peanut instead, which caused the very same cell in the monkey s brain to fire. What makes these neurons special is that there is no distinction between the monkey s own performance of an action and the mere observation of another performing the very same action (De Waal, 2009). This discovery has been said to be as important to psychology as the discovery of DNA to biology, and the fact that this was done in monkeys further shows that empathy is not unique to humans (De Waal,

10 EMPATHY FOR PAIN ). However, the existence of mirror neurons is not proven to be the same as empathy, and the role they play in empathy is not clear yet (Lamm & Majdandžić, 2015). There are differing opinions among scientists as to whether mirror neurons correlate with action understanding or if they are simply a part of the action. In the same way, the activity in shared networks involved with empathy for pain could either be thought of as a route to understanding the feelings of others or as a sign of it (Lamm & Majdandžić, 2015). The hypothesis that mirror neurons reflect action understanding has been heavily questioned and Hickok (2009) puts forward arguments against this proposal. He states that a motor representation cannot distinguish between the range of possible meanings associated with such an action (Hickok, 2009, p. 1240). This means that an action consists of several elements and that the intentions and components of an action could be many. Therefore, because of this range of possible meanings and ways to achieve a goal, there has to be a clear distinction between the goal and the specific motor actions necessary to achieve it. The hypothesis that action understanding is a consequence of a similar activity in our own brain would then be false, either because mirror neurons do not code actions, or because motor representations are not the basis of action understanding (Hickok, 2009). The brain areas that are associated with mirror neurons are collectively called the mirror system (Baron-Cohen, 2011). As mentioned earlier, these areas are active when an individual performs an action as well as when observing someone else performing the very same action. For ethical reasons, it has been somewhat difficult to establish which areas the human mirror system might contain, although it has been suggested that it involves the IFG, the inferior parietal lobule (IPL) and the inferior parietal sulcus (IPS) Although there is no direct evidence that mirror neurons reflect empathy (Lamm & Majdandžić, 2015), the mirror system is implicated in mimicry and emotion contagion. For example, when someone else

11 EMPATHY FOR PAIN 11 yawns and one involuntarily yawns too, these areas are activated. Actions such as these, which are called mimetic actions, occur automatically (Baron-Cohen, 2011). The empathy circuit. The neuroscientific understanding of empathy has increased quickly, thanks to a growing number of studies (mostly carried out with fmri because of its availability and precision) which have enabled us to associate different brain regions with components of empathy (Lamm, & Majdandžić, 2015). Frequently, the brain areas involved with empathy have been referred to as the empathy circuit (e.g. Baron-Cohen, 2011) The medial prefrontal cortex (MPFC) is referred to as a centrality that processes social information. It is also involved with the ability to compare one s own perspective with another s. The dorsal part of this region (dmpfc) is associated with thinking about the thoughts and feelings of other people, as well as our own. The ventral part (vmpfc) is on the other hand involved with merely the thoughts of the self and one s own mind, rather than someone else s (Baron-Cohen, 2011). It has also been hypothesized by Damasio that this area has to do with the processing of emotional valence in actions. Damasio proposes a so called somatic marker, which is the emotional outcome of an action and the thing that makes us tend to repeat only the actions that we associate with positive emotions (Damasio, Everitt, & Bishop, 1996). Also, this theory is supported by findings demonstrating that the vmpfc is one of the regions involved with the control of mood-related behaviors (Lim, Janssen, Kocabicak, & Temel, 2015). The vmpfc is one of the regions that show abnormally low activation in people with low empathy (Baron-Cohen, 2011) and it also seems to be involved with cognitive empathy (Shamay-Tsoory, Aharon-Peretz, & Perry, 2009). The orbitofrontal cortex (OFC) is a part of the vmpfc. When OFC is damaged, patients lose their social judgement which affects the social behavior. Further, OFC

12 EMPATHY FOR PAIN 12 impairments are often associated with empathy dysfunction and psychopathy (e.g., Baron- Cohen, 2011; Shamay-Tsoory et al., 2009). Adjacent to the OFC is the frontal operculum (FO). The adjacent part of the FO is, together with the anterior insula, involved with the processing of other s negative emotions and negative experiences such as pain. They are also associated with positive feelings of another, in that these two areas map the bodily feelings of others into the internal state of the observer (Jabbi, Swart, & Keysers, 2007). Also, the posterior insula has been suggested to support an early convergence of sensory and affective processing (Morrison, Löken, & Olausson, 2010). Moving on to the area located inferior to the FO, called the inferior frontal gyrus (IFG), this area is important in the processing of emotional faces (Baron-Cohen, 2011). Moreover, the IFG is involved with affective empathy and if damaged, this capacity is heavily disturbed (Shamay-Tsoory et al., 2009). The middle cingulate cortex (MCC) plays a role in empathy in that it is associated with bodily aspects of self-awareness (Baron-Cohen, 2011). Together with the anterior insula (AI) it is activated during the experience of pain as well as when observing others in pain (Baron-Cohen, 2011; Singer et al., 2004). Damage to these regions can disrupt the ability to recognize emotions such as happiness, disgust and pain (Baron-Cohen, 2011). The temporo-parietal junction (TPJ) seems to be important to ToM, in that it is associated with the judgement of other s intentions and beliefs (Saxe & Kanwisher, 2003). It is also involved with the self-other distinction (Schulte-Rüther et al., 2008). Superior temporal sulcus (STS) is associated to empathy because of animal studies revealing neurons in STS have been found to respond when the animal observes someone else s gaze (Baron-Cohen, 2011). Further, this area seems to be involved with the

13 EMPATHY FOR PAIN 13 processing of biological motion and facial expressions of others. The STS might also play a role in emotional perspective taking. (Schulte-Rüther, Markowitsch, Shah, Fink & Piefke, 2008). The last brain structure that has been concluded to be a part of the empathy circuit is the amygdala (Baron-Cohen, 2011). The amygdala is associated with emotional learning and regulation processing (Wager, Davidson, Hughes, Lindquist, & Ochsner, 2008). It is also involved with empathic regulation towards the emotional pain of others such as deliberate control of self-focused distress (Bruneau et al., 2015) The abovementioned areas are the major structures implicated in empathy (Baron-Cohen, 2011) and moreover the most important regions to this thesis. In the following, it will be discussed how their activity may be represented in the observer of a person suffering from pain, as well as in the suffering individual. Keeping our current knowledge of empathy in mind, some differences between these activities and the consequences of how impairments to some of these areas may affect empathy will be presented. Pain Theories of Pain Physical pain includes a number of events such as tissue damage, visceral unpleasantness, arousal, a change in the direction of attention, negative affect as well as a desire to withdraw from similar experiences. Most of these features are nonspecific when looked at individually, in that they do not occur merely as a reaction to pain. Therefore, nociceptive pain does not represent pain as such; rather, it representsthe combination of several features (Zaki, Wager, Singer, Keysers, & Gazzola, 2016). This combination of events is referred to as nociceptive pain, which is the first hand-experience of physical pain (Zaki et al., 2016). The term derives from nociceptors; the kind of receptors that are activated when

14 EMPATHY FOR PAIN 14 non-neural tissue is damaged. In the skin, there are different nerve fiber endings, which possess different functions. These fibers are axons of cells and neurons that project to the dorsal root ganglia (DRG) or the trigeminal ganglia (TG) (Chuquilin, Alghalith, & Fernandez, 2016). The different types of fibers are categorized by means of their diameter and myelination. It is the small fibers with little or no myelination that are responsible for the sensation of pain, and that are called nociceptors (Chuquilin et al., 2016). C-fibers, which are unmyelinated and slowly conducting, are often polymodal and respond to noxious thermal mechanical stimuli (Julius & Basbaum, 2001). Also, many C-fibers respond to chemical noxious stimuli such as acid or capsaicin. Aδ-fibers are, on the other hand, slightly myelinated and therefore conduct faster than C-fibers (Weiss et al., 2008). They respond to intense mechanical stimuli, and can alternate in how they are affected by tissue damage or heat (Julius & Basbaum, 2001). It has been assumed that Aδ-fibers mediate so called first pain, which is rapid, acute and sharp pain, whereas C-fibers are thought to mediate second pain, namely the delayed, more diffuse and dull pain (Julius & Basbaum, 2001). In relation to selective noxious stimulation of tiny parts of the skin, C-fiber stimulation was strongly associated with activity in the right frontal operculum and the anterior insula, whereas stimulation of Aδ-fibers was not. Based on current knowledge of these structures, it has been further suggested (Weiss et al., 2008) that C-fibers might be engaged in homeostatic and interoceptive functions in another way than Aδ-fibers, producing a signal of greater emotional salience (p. 1372). Information from the nociceptors travels to different parts of the brain via different pathways from the spinal cord (Apkarian, Bushnell, & Schweinhardt, 2013). Historically, the study of pain has resulted in differing theories about its origins and characteristics. Specificity theory stated that pain is a specific modality, such as vision or hearing. According to this theory, pain receptors in the body tissue project to a specific center

15 EMPATHY FOR PAIN 15 in the brain, where it is processed (Melzack & Wall, 1996). However, it has been argued that clinical, physiological and psychological experiences do not support the idea of a direct transmission from stimulus and sensation to the brain (Melzack & Wall, 1967). For example, there is evidence from patients and cases in which the subject experiences severe pain even though the pain stimulus is mild, as well as when subjects do not experience pain at all despite powerful painful stimuli. Moreover, there have been difficulties in locating fibers that always respond to harmful stimuli (Melzack & Wall, 1967). As a reaction to specificity theory, a number of alternative theories came to existence, which can be grouped under the term pattern theory. These pattern theories suggested that there are no specific pain receptors; rather, the experience of pain arises from an intense stimulation of nonspecific receptors, which in turn creates a nerve impulse pattern (Melzack & Wall, 1996). However, most of the pattern theories failed to acknowledge the existence of highly specialized fibers and receptors (Melzack & Wall, 1967). The proposal of the so-called gate control theory of pain marked a significant change in thinking and theorizing about pain Gate control theory of pain. The gate control theory of pain by Melzack and Wall (1967; 1996) contributed to the understanding of pain by its emphasis on central neural mechanisms. It forced the medical and biological sciences to consider the brain as an active system that modulates inputs (Melzack & Katz, 2004). The theory suggests a gating mechanism which mediated modulation of pain information and this proposal received greatreception and research emerged to support or disprove it. The human spinal cord includes gray matter that forms three pair of horns; the dorsal, lateral and ventral horns. Out of these three, it is the dorsal horns that have been shown to be important for the processing of pain in that they receive information from primary

16 EMPATHY FOR PAIN 16 afferent axons such as nociceptors that respond to for example tissue-damaging stimuli from the skin, muscle joints and viscera (Todd & Koerber, 2013). The gate control theory states that inhibitory interneurons in the dorsal horn are important to the control of incoming sensory information and are thereafter sent to the brain (Todd & Koerber, 2013). The foundation of the gate control theory rests upon the fact that impulses from peripheral stimulation are transmitted to the following three spinal cord systems: substantia gelatinosa fibers in the dorsal horn; dorsal column fibers that project to the brain; and the so called T cells in the dorsal horn (Melzack & Wall, 1967). The substantia gelatinosa consists of small and densely packed cells that together form an extension of the spinal cord. In the dorsal horn, nerve impulses from peripheral fibers are thought to be modulated. Thus, the substantia gelatinosa functions as the gate control system. The modulation of the afferent patterns further activates T cells, which in turn gives rise to the activation of neural mechanisms which comprise an action system. In this system, the perception of the painful stimuli is processed, as well as the response to it. The gate control system in the substantia gelatinosa is affected by the so called central control trigger which contains of the afferent patterns in the dorsal column. The central control trigger causes the activation of selective brain processes, influencing how the nerve impulses will be modulated within the gate control system (Melzack & Wall, 1967). The gate control theory suggests that the experience of pain is a consequence of an interaction between the three abovementioned systems (Melzack & Wall, 1967). Since the theory was proposed the role of the dorsal horn has been intensively studied; however, there is limited knowledge within this field. What is known is that the dorsal horn consists of four neuronal components; primary afferent axons, interneurons, projection neurons and descending axons. Altogether, these components make up for the response to and modulation of sensory and nociceptive information and the transmission of this information to various

17 EMPATHY FOR PAIN 17 parts of the brain and other spinal segments (Todd & Koerber, 2013). The next section will focus on pain and how the impulse projects from the onset of harmful stimuli to the brain, giving rise to the overall experience of pain. Pain Pathways The multidimensional experience of pain implies different pathways involving different structures in the brain. In order to clarify the broadness of pain, we need an understanding of these pathways and how they comprise several brain areas, resulting in the many components of the pain sensation. The pathway most commonly associated with pain (Dostrovsky & Craig, 2013) is called the spinothalamic projection because pain related information is sent directly from the spinal column to the thalamus. The information can also be sent from the thalamus to homeostatic control regions in the medulla and brain stem. Spinal input to the brain stem affects the spinal and forebrain activity, which in turn might influence pain experience (Dostrovsky & Craig, 2013). When the information is sent to the brain stem it is called spinobulbar projection. This pathway is involved with the integration of nociceptive activity with processes associated with homeostasis and behavior. However, direct projections to the hypothalamus and ventral forebrain are also possible. In addition to the spinothalamic and spinobulbar projections, which are considered to be the two most prominent pathways, there are also indirect pathways that seem to be related to pain. These are the post-synaptic dorsal column (PSDC) system and the spinocervicothalamic (SCT) pathway. The PSDC and SCT pathways both originate from cells in the spinal dorsal horn and signals project from the brain stem to the forebrain (Dostrovsky & Craig., 2013).

18 EMPATHY FOR PAIN 18 The discovery of the abovementioned pathways expands our knowledge about the multidimensionality of pain because it shows that the experience of pain is due to several factors and projections and relates to different neural structures. However, it is important to keep in mind that this experience is due to the activity in many brain regions simultaneously. As Dostrovsky and Craig state, it is the constellation of activity across the entire brain that must constitute the basis for the conscious experience of pain. (Dostrovsky & Craig, 2013, p. 196). Neural Correlates of Pain Information from nociceptors projects via different pathways to the brain, mostly through the thalamus, which is interconnected to different structures within the cerebral cortex (Dostrovsky & Craig, 2013). When talking about the neural correlates of pain one often refers to the involved brain structures as a pain matrix (Baron-Cohen, 2011; Singer et al., 2004). The pain matrix consists of the following areas: the primary somatosensory cortex (SI), the secondary somatosensory cortex (SII), insular regions, the anterior cingulate cortex (ACC), the movement-related areas such as the cerebellum and supplementary motor areas and the thalamus (Singer et al., 2004). The ACC and the MCC are associated with the affective-motivational aspects of pain and also with functions linked to emotional experiences (Lamm & Majdandžić, 2015). Especially important to the study of pain is the thalamus. The lateral thalamus is associated with discriminative pain, whereas the medial thalamus is involved with the motivational aspects of pain. However, it is important to keep in mind that the sensation of pain does not occur within the thalamus, but its interconnection to the cerebral cortex gives rise to this particular experience (Dostrovsky & Craig, 2013). As noted earlier, a growing number of studies suggest that a consistent cortical and subcortical network has been found to respond to pain in healthy subjects. Commonly, the

19 EMPATHY FOR PAIN 19 areas involved with this network are the SI, SII, ACC, IC, PFC, thalamus and cerebellum. It seems that the somatosensory cortices are associated with the perception of sensory features such as location and duration of pain, and the limbic and paralimbic regions such as the ACC and IC are associated with the emotional and motivational aspects of pain (Apkarian et al., 2013). Empathy for Pain Neural Correlates of Empathy for Pain When one is experiencing empathy for another s pain, this involves a complex process in which the sensory and emotional qualities of vicarious pain are extracted and then mapped onto neural substrates which may or may not be those that are activated in the direct experience of pain (Betti & Aglioti, 2016). However, some studies indicate that it is only the affective component of the pain matrix that is involved with empathy for pain (Singer et al., 2004) which suggests that it is only the emotional representation of pain that is shared between the observer and the sufferer (Avenanti, Paluello, Bufalari, & Aglioti, 2006). One of the first studies with focus on the social neuroscientific study of empathy for pain was carried out by Singer et al. (2004). By using fmri, the authors investigated which brain regions were active when participants experienced pain, as well as when observing another person suffering from pain. This provided evidence for pain-related responses and also confirmed that empathic experience does not involve activation of the entire pain matrix, but only the regions associated with the affective dimension of the experience of pain. The study was carried out with the help of 16 couples, based on the assumption that couples are likely to empathize with each other. The authors scanned the female partners with fmri and exposed either them, or their partners, to pain using an electrode attached to the back of their right hands (Singer et al., 2004). A mirror was placed so that the female could see both her and her partner s hand. Stimulation was either low (no

20 EMPATHY FOR PAIN 20 pain) or high (painful) and presenter to either the female or the male. When females were presented to pain ( self condition), there was an increased activity in the following brain areas: contralateral SI/MI, bilateral SII with peak activation in contralateral posterior insula extending into SII, bilateral mid and anterior insula, ACC, right ventrolateral and mediodorsal thalamus, brainstem, and mid and right lateral cerebellum (Singer et al., 2004). These regions, which make up for the pain matrix have been associated with painful stimuli in several studies (Bruneau et al., 2015; Jacoby, Bruneau, Koster-Hale, & Saxe, 2016; Lamm, Decety, & Singer, 2011). When the male partners were presented with painful stimuli ( other condition), the following brain regions showed activation in the female s brains: ACC (more specifically, the anterior and posterior rostral zones), the AI bilaterally with an extension into inferior prefrontal cortex, cerebellum and the brainstem. These activations show that some of the areas in the pain matrix are activated also when observing another person suffering from pain. However, the absence of activity in the somatosensory cortex is one distinct factor that differentiates empathic pain from physical pain, in which there is activity in this region. Also, in the other condition, there were observed activations in the ventral and dorsal visual stream, including bilateral fusiform cortex, lateral occipital and right posterior superior temporal sulcus, the left inferior parietal cortex, and the left superior frontal cortex. (Singer et al., 2004) Results from this study showed that there was an overlapping neural activation in cingulate and insular cortices during the direct experience of pain as well as when empathizing with the pain of others. In other words, empathy recruits similar neural networks as the direct experience of the emotion one is showing empathy for (Singer et al., 2004). Because of these similarities between the neural activations for self-and-other-related experiences, it has been suggested that the ability to experience empathy is partially based upon the processing of our own emotions (Kanske et al., 2015).

21 EMPATHY FOR PAIN 21 Shared networks. As noted earlier, empathizing with another person s feelings has been shown to activate the neural networks that are also related to the direct experience of the very same feeling (Betti & Aglioti, 2016). In order to study empathy on a neuroscientific level, researchers often study these shared neural networks in relation to pain. This is mostly done by either presenting participants with painful stimulation to their bodies, or by presenting them with pictures or cues that indicates that another person experiences pain. The most common way to measure the brain activity in such studies is by using fmri. Thereafter, comparisons are made between the first-hand experience of pain and the observation of another person suffering from pain. These kinds of studies have repeatedly shown that shared neuronal networks exist (Singer & Klimecki, 2014). When empathizing with another individual suffering from pain, we talk about empathic pain (Zaki et al., 2016). The overlap between nociceptive and empathic pain has been noticed and investigated for decades and research has shown that specific brain structures such as the anterior insula (AI) and parts of the cingulate cortex (CC) are typically activated both in the experience of nociceptive pain and empathic pain (e.g., Lamm et al., 2011; Singer et al., 2004). However, this activation of AI and CC might be represented in other psychological states as well, such as attention or arousal (Singer, Critchley, & Preuschoff, 2009). Hence, there is currently a debate among researchers on whether the similarities between nociceptive pain and empathic pain suggest shared pain-specific processes or if these findings are a consequence of incorrect reverse inference (Zaki et al., 2016). This would mean that neural activity is incorrectly associated with a certain cognitive process. Even if pain is the most widely used and common way to study empathy, similar studies have also shown similar paradigms of a shared network in field of touch, disgust, taste and social rewards. A shared network has been observed in the somatosensory cortex in

22 EMPATHY FOR PAIN 22 relation to vicarious neutral touch, and in the medial orbitofrontal cortex in relation to vicarious pleasant touch. In the study of shared social rewards, a shared network has been found in the ventral striatum, and in the study of taste and disgust there is a shared network in parts of the AI (Singer & Klimecki, 2014). Somatosensory system. In the experience of empathy for pain, both affective and sensorimotor pathways are involved. However, the question of what role the somatosensory cortex (SI) plays in this process remains unanswered (Betti & Aglioti, 2016). For example, experimental paradigms may affect the activation in the somatosensory cortex since it is involved with touch and some experiments show visual cues where models are being touched (Baron-Cohen, 2011; Lamm et al., 2011). Also, experiments in which participants are presented to abstract visual cues that indicate the pain of another, did not show significant activation of SI and SII (Singer et al., 2004). Although SI is mostly considered to be involved with somatic processing, it also seems to play an important role in complex cognitive functions such as social cognition. Activity in the somatosensory structures during observation of the emotional state of others could provide the observer with a somatic reflection of what that particular emotional state may feel like (Bufalari, Aprile, Avenanti, Di Russo, & Aglioti, 2007). Bufalari et al. (2007) used somatosensory-evoked potentials (SEPs) together with EEG to investigate if observing a model suffering from pain or tactile stimuli modulates neural activity in the somatic system of the observer. SEPs are a noninvasive, non-painful way to assess somatosensory system functioning and were in this experiment obtained by electrical stimulation of the right median nerve at the participants wrists. Attendants were presented to video clips where models were experiencing pain or tactile stimuli. Activations in the primary somatosensory cortex (SI) correlated with the intensity but not the unpleasantness of the pain and touch. The results from this study suggest that neural activity

23 EMPATHY FOR PAIN 23 in SI is not only involved with the actual experience of pain and touch, but it also seems to be modulated by the observation of others bodily sensations (Bufalari et al., 2007). Factors Influencing Empathy for Pain There are several factors that can influence to what degree one feels empathy for another individual s pain, such as the affective link or similarity between observer and sufferer (Loggia, Mogil, & Bushnell, 2008), the age (Bandstra, Chambers, McGrath, & Moore, 2011) and gender (Christov-Moore et al., 2014) of the observer and the context in which the empathy is experienced. Factors such as these will be explained and discussed in terms of behavioral and neural mechanisms in this following section. Age. Already by the age of 5-6 years, it is very clear that children are able to recognize and identify pain in others, and refinements of this ability continue through to early adulthood (Deyo, Prkachin, & Mercer, 2004). However, little is known about how and when children develop or express empathy for another individual s pain (Bandstra et al., 2011). The empathic response first emerges as a reaction of personal distress and is often accompanied with self-focused behavior such as self-soothing. As the child matures, the control of the own emotions typically becomes better and the focus on the needs of an individual in distress becomes greater (Bandstra et al., 2011). In a test where 120 children between the ages of 18 and 36 months were presented with simulations of an adult s pain and sadness respectively, their empathy-related behaviors were investigated. The results in this study indicated that children tended to be more sensitive to others sadness, in which they became distressed and showed prosocial behavior. When presented with others pain, the children were more likely to continue playing. However, behaviors associated with empathic concern and personal distress emerged in both situations. It was speculated that the reason for a reduced reaction to the pain stimulus might be a consequence of the children not regarding the pain of others as

24 EMPATHY FOR PAIN 24 threatening to their selves. Further, this finding has been suggested to be due to the frequency of painful events that occur in childhood, thus leading the child to become inured to observing others in pain (Bandstra et al., 2011). Older children were more likely to react to the pain stimulus with empathic concern and less likely to react with personal distress than younger children. In the sadness condition, this age difference was however not present, suggesting that the developmental course of empathy for pain and empathy for sadness are different (Bandstra et al., 2011). In one study investigating the development of empathy, fmri was used to scan and compare children and adults (Decety & Michalska, 2010). Participants in the age of 7-40 years old were presented with animations depicting painful and non-painful conditions. Results from this study showed that children and adults have similar patterns of brain activity when observing other people in pain: in the ACC, somatosensory cortex, and periaqueductal gray (PAG) and the insula. Despite this, there are some important age differences in the neural activity associated with empathic pain. For example, the amygdala and the posterior insula are developed much earlier in childhood than other structures such as the dorsal and lateral vmpfc, which are slower to mature and later on become specialized for the evaluation of social stimuli (Decety & Michalska, 2010). Also, different parts of the prefrontal cortex (PFC) mature at different rates, indicating that the development of affective processing from childhood to adulthood is associated with reduced activity in the limbic affect processing systems, whereas there is an increased activity in other prefrontal systems (Decety & Michalska, 2010). Further, older adults seem to be more sensitive than younger people to the intentional harm of others (Chen, Chen, Decety, & Cheng, 2014). The empathic response in the mpfc and STS does not change with age, although, the activity in the AI and anterior mid-cingulate cortex in response to others pain has been shown to decline with age. This

25 EMPATHY FOR PAIN 25 finding indicates that the neural response associated with affective empathy lessens with age, whereas the response to perceived agency is preserved (Chen et al., 2014). Gender. Sex differences in empathy-like behaviors have been reported in nonhuman animals such as primates and rodents and suggest that females possess greater levels of empathy in at least some species (Christov-Moore et al., 2014). It has been shown that there are human gender differences in empathy for pain, in that both the short-latency empathic response and the long-latency empathic response to a stimuli depicting someone suffering from pain differs (Han, Fan, & Mao, 2008). There also seems to be a gender difference in the activation of the MCC and the AI in relation to observing others in pain. When men observe someone they do not like or consider fair, they generally show less activity in these areas than when they like the suffering person (Singer et al., 2006). Females have been shown to have higher emotional responsiveness and mirroring responses to the suffering of others, and they are also better at recognizing emotions (Christov-Moore et al., 2014). Further, it has been suggested that females tend to show more prosocial and altruistic behavior. In other words, females seem to be show higher affective empathy for other s pain. In contrast, in terms of cognitive empathy, it has been suggested that males show more utilitarian behavior than females, accompanied by a greater recruitment of areas that are associated with cognitive control and cognition (Christov-Moore et al., 2014). In a study where gender differences in the neural networks associated with empathy were investigated (Schulte-Rüther, Markowitsch, Shah, Fink, & Piefke, 2008), fmri was used to scan subjects while they were presented with different emotion expressing faces. The participants were asked to either focus on their own emotional response to the presented picture, or to evaluate the emotional state that the observed face expressed. Females rated

26 EMPATHY FOR PAIN 26 their own response higher than males. The females brain activity in this task involved a stronger activation in the right inferior frontal cortex and STS whereas the males showed a stronger activity in the left TPJ. Since the TPJ is associated with the self-other distinction, this suggests that males rely more on this distinction than females. In the task were the participants were asked to evaluate the observed face s emotional state, females showed an increased activity of the right inferior frontal cortex in contrast to males. Taken together, results from this study suggest that females recruit more brain areas containing mirror neurons than males do and that these gender differences may indicate different strategies in how own emotions are processed in response to others (Schulte-Rüther et al., 2008). Empathy for non-human entities. The ability for humans to empathize with other animals in a similar way as with their own offspring has been suggested to depend on an instinct of nurturance (Prguda & Neumann, 2014). Using fmri, Mathur, Cheon, Harada, Scimeca and Chiao (2016) investigated whether the neural reactivity associated with empathy for pain was different when directed to humans than to non-human entities. They measured neural activity when subjects were presented with visual scenes depicting people, animals and nature in either negative or neutral conditions. The same brain regions that increased in activity while subjects were experiencing empathy for other people were active when they observed animals and nature in harmful conditions (e.g. dorsal anterior cingulate cortex, bilateral anterior insula). These findings suggest that the activity within these areas is not specific to humans only, but also when empathizing with other animals and the nature (Mathur et al., 2016). However, empathy for humans and other animals has shown to be facilitated by the perceived phylogenetic similarity between the object and subject (Prguda & Neumann, 2014). Stronger phasic skin conductance response (SCR) and subjective ratings of empathic experience have been associated with human participants observing the suffering of phylogenetically similar species than more different animals. The subjective ratings did not

27 EMPATHY FOR PAIN 27 differ much when subjects were presented to suffering humans, non-human primates and quadruped mammals. On the other hand, the arousal and subjective experience of empathic emotions were rated as much lower when observing birds suffering from pain (Prguda & Neumann, 2014). Also, SCR showed that participants were more emotionally aroused and directed more attention toward stimuli depicting suffering humans than to non-human primates. This response was further decreased when participants were presented to stimuli of suffering quadruped mammals, and even lower in the bird stimuli. This supports the idea that phylogenetic similarity plays an important role in empathy for the pain of others (Prguda & Neumann, 2014). Empathy among caregivers. In medical practice, pain is most commonly defined by its pathological cause. In cases where a physical cause is not found, the patients pain is often thought to be imaginary (Ojala et al., 2015). In order to investigate invisible chronic pain, Ojala et al. (2015) studied the contact between patients and care givers. The patients report that their experience of pain is often underrated or denied. Further, many of these patients have problems trusting and believing in their helpers and they also have mental health problems together with the painful experience. Even though there is a current recommendation to treat chronic pain as a biopsychosocial experience, it is often thought to be a symptom of an underlying disease (Ojala et al., 2015). In other words, too little empathy among caregivers may have devastating effects on the health of patients. However, failure to help patients suffering from chronic pain may also have negative effect on the caregivers. Too much empathy can cause the helper to suffer more than usual with the patient, which can result in a desire to withdraw from this kind of work. Caregivers with too much empathy tend to continue with treatments even if they are not helping, or even worse, if they risk damaging the patient (Breivik, 2015).